464 Dr. B. Stewart on the Kew Magnetic Curves. [June 10, 



Nevertheless one notable exception to this rule was mentioned in the 

 above paper, namely, that at the beginning of the great disturbance of 

 August-September 1859, an abrupt fall of the declination curve corre- 

 sponded to a rise of the other two components. It was also shown in 

 this paper that while the horizontal-force peaks are always as nearlv as 

 possible double in size of the vertical -force peaks, the proportion between 

 the declination peaks and those of the other components appeared to be 

 variable. Some light was thrown upon this variability in a subsequent 

 paper by Senhor Capello and myself, in which the peaks and hollows at 

 Lisbon and at Kew were compared together (Proc. Roy. Soc. 1864, p. 111). 

 It was found that these phenomena occurred simultaneously at these two 

 observatories ; and it was stated that, as far as Kew is concerned, the pro- 

 portion of the declination peaks and hollows to those of the horizontal 

 and vertical force presents the appearance of a daily range, being great at 

 the early morning hours and small in those of the afternoon. 



Thus the type of small and abrupt changes, judging from the behaviour 

 of the declination, seemed to vary from two causes, being in the first place 

 subject to a diurnal variation, and in the second place appearing to vary 

 with the disturbance, inasmuch as that for the great disturbance August- 

 September 1859 was, as above stated, entirely different from the usual type. 



This complexity seems puzzling ; but the results of a preliminary com- 

 parison between the Stonyhurst and Kew declination magnetographs 

 (Sidgreaves and Stewart, Proc. Roy. Soc. 1869, p. 236) appear to throw 

 some light upon its cause. It was there stated that when the declination- 

 curves of Stonyhurst and Kew are compared together during rather slow 

 disturbances, the scales are such that the traces seem exactly to coincide 

 even to their most minute features ; but, on the other hand, when the 

 disturbance is abrupt, there is an excess of Stonyhurst over Kew, which 

 appears to vary with the abruptness of the disturbance, being great when 

 this is great. In fine, there appears to be superimposed upon a disturb- 

 ance, which is mainly cosmical, a comparatively small effect, which appears 

 to be of a more local nature, and may perhaps be caused by earth-currents. 

 This circumstance renders it prudent, in discussing the laws of the small 

 and abrupt changes of force (peaks and hollows) at Kew, to avoid all great 

 and excessively abrupt disturbances, confining ourselves to those cases in 

 which there is only a moderate abruptness. The result obtained for the 

 great disturbance August-September 1859 may therefore be dismissed 

 as probably effected by this local cause, inasmuch as the disturbance 

 measured was very abrupt. The question then arises — Rejecting very 

 great and abrupt disturbances, has the peak-and-hollow force only a regular 

 diurnal variation, or is it subject besides to other changes of type ? 



Mr. Whipple, magnetical assistant at Kew, has carefully selected and 

 measured all the similar peaks and hollows for the first two years of the 

 Kew curves ; and the result exhibits a manifest diurnal variation in the 

 type of the peak-and-hollow force. 



